Content, context and predicting viral spread
Molly Flatt from 1000heads pointed towards an article in New Scientist lately. Spanish researchers José Luis Iribarren from IBM and Esteban Moro at the Carlos III University in Madrid developed a method to accurately predict how quickly and widely new pieces of information, or “memes“, will spread.
“The secret, they say, is to recognise the fact that people vary in how “infectious” they are when it comes to sharing content online. While some people pass on things they receive right away, others do so after some delay, or not at all.”
Basically, they combine the basic reproductive number, R0, with some expectation of the variation in the time it takes for people to respond to a meme ( i.e. pass it along). They had found there was a huge variation at the individual level.
Combining these two numbers, the researchers could build a model to predict the meme’s spread.
“The model cannot predict whether a piece of content will go viral before it has been released; only its likely reach once it starts spreading.”
The fact that individuals differ in whether and how fast they pass things on is hardly news. But the ability to predict the final results/reach of a campaign is no doubt interesting.
Also, it is interesting to consider the differences between a campaign which might attract a “slower” audience, creating a longer and thinner peak in the statistics, compared to a more classic viral that peaks rather quickly and then enters a kind of “long tail” phase.
Two final notes:
First, in the article, David Liben-Nowell notes that: “…people may vary in infectiousness depending on the type of content they receive.” This is highly dependant on how well targeted the content is to them. If they are attracted by it, the likelihood they will spread it is higher and I assume they will also pass it on quicker.
Second, just as Duncan Watts says, context, or how receptive a society as a whole is to an idea, is extremely important. I assume this too will affect if people will pass something on, as well as how quick they are to push the “send” button.